Page 1 of 4 1234 LastLast
Results 1 to 20 of 65
  1. mslade's Avatar
    Posts
    16 Posts
    Global Posts
    18 Global Posts
       #1  
    I just spoke to Palm support and they admit there is a 'battery problem' with new treo 680s. They claim that engineering is working on the 'fixing the battery' but the date for a battery replacement is not known.

    However, I don't think they understand the problem!

    I have just installed BatteryGraph and this seems to indicate that there is a firmware problem based on the following analysis.

    Immediately after a full charge, BatteryGraph shows full percentage charge ans well as full mv scale - in my case 3960mv. With no use, the percentage drops off precipitously - within 24 hours down to under 20%! This is similar to what the battery icon shows.

    However, the BatteryGraph MV reading is still at 3740mv! This is in the expected range.

    With the icon reading governing low battery warnings, etc. everyone thinks they have a battery problem but the real problem is that the firmware isn't calculating usage right from the mv reading! Since charging seems to follow the battery icon, the problem is likely to result in overcharging, too.

    Please try using BatteryGraph and report your results! I found it through mytreo.net. This is the same tool that someone else has used to confirm that the talk time is about as claimed. However, this person did not do a standby time analysis.

    If you have any ideas on how to spread the word about this test and/or a way to get this information to the right people at Palm, please do so.

    thanks
  2. jfme's Avatar
    Posts
    79 Posts
    Global Posts
    82 Global Posts
    #2  
    So this is a problem with the battery meter algorithm.

    No wonder why some have reported up to 90 minutes usage after showing 0% capacity.
  3. mslade's Avatar
    Posts
    16 Posts
    Global Posts
    18 Global Posts
       #3  
    That is my analysis but I'd like to see some confirmation-

    -that others get the same results
    -analysis that confirms battery meter algorithm

    And more important-some way of getting the suggestion to Palm engineering!
  4. #4  
    I think you are a little confused as to how a lithium ion battery works.
    3740-3760mV is correct as this translates to the 3.7V rating of the battery.. Lithium ion batteries provide the same voltage until it is almost dead, as opposed to NiCD batteries which drop voltage as capacity decreases.
    The Amperage is what the percentage battery left reports.
    Voltage of the battery will stay the same until the very end of battery life. The amperage will decrease throughout the day.
  5. #5  
    Anybody also notice how warm the back of the 680 phone is when you are using it? Never noticed that with my 650...and since the warmth is right over the battery area, same source of all my headaches with this new Palm, I wasn't sure if this observation is significant so wanted to mention it in case Palm is reading these!
    Last edited by Rebeccaone; 01/05/2007 at 12:21 PM. Reason: (typo in original post)
  6. #6  
    I've noticed warmth over the battery compartment with both my old 270, 600, and 650 and my new 680. I don't think it is a new phenomenon.
  7. mslade's Avatar
    Posts
    16 Posts
    Global Posts
    18 Global Posts
       #7  
    My battery icon now reads '0'. Message comes up 'Battery is extremely low'.

    BatteryGraph reads 3670 mv, 3%. Once again this suggests firmware problem!

    Using a voltmeter on the battery it confirms 3.67 volts.

    Even given the non-linear charge vs. voltage curve of this type of battery, it looks like an algorithm problem.
  8. aaronchow's Avatar
    Posts
    51 Posts
    Global Posts
    94 Global Posts
    #8  
    mslade,

    Calm down, please read the above post written by howiecam.
  9. #9  
    Also, once you have done this a time or two, your battery charging circuit will get calibrated and you will not be able to get to 0%. When I get low battery warnings now, I'm still in the teens. The treo 680 now shuts off well above 0%. In short, my charging circuit has calibrated.
  10. #10  
    Frankly, it's interesting to postulate as to what the problem is BUT really, I just want the solution. In this case, I believe it's the deep drain and recharge to help calibrate your battery meter.

    The cause and software fixes I'll leave to palm, if they ever get around to it.
  11. mslade's Avatar
    Posts
    16 Posts
    Global Posts
    18 Global Posts
       #11  
    I've read it.

    I'm calm but annoyed.

    I'm a retired physicist.

    I've read some of the papers on measuring Li ion battery charge capacity. It is not an easy problem to measure.

    But 3.67+ volts does not seem to square with less than 20% capacity remaining.

    All I am saying is other people should look at the discharge information that the BatteryGraph plugin yields for their Treo 680. Are the results consistent? Do they supply anything useful to understand why battery charge lifetime is so short on the 680?

    Since the data using the MV readings have been used to suggest that the talk time spec is reasonable, the same tool may give information on the standby time/no phone on at at all charge time.

    No one has provided any data that the 12 day standby spec can be confirmed under any conditions.

    The talktime data that I have seen using Batterygraph approximately confirms the specification. Therefore I conclude that it is not a battery capacity problem.

    Since turning the Treo 680 phone off and no use at at should give much longer than 12 days (Bluetooth off, etc.) but apparently does not, it is unlikely that the problem is understood to be an undercapacity battery. It might be a battery that just doesn't hold charge. (I've had mine replace once. Same problem.)

    Given that the mv discharge curve looks reasonable for 12 days standby and the percentage curve does not, this information that suggests that it is a firmware rather than a battery problem.

    Please try BatteryGraph or some other charge capacity software and supply some data.
  12. #12  
    I used BatteryTime after the first deep discharge/charge cycle (with soft reset in between while connected to charger, this recalibrates the charging circuitry).
    BatteryTime is an application which continuously logs to memo time elapsed and battery level (have to use AlwaysOn as well or it will shut down after the system set time-out).
    I did not run the test until the battery was flat (had to stop around 30%) but estimated end result would have been close to 7 hours.
    This was with GSM radio ON !!

    Have you already tried to recalibrate the battery circuitry by connecting to charger, remove the battery, wait some seconds, replace the battery and let it charge for at least 3 to 4 hours ?
    Are you also aware a new LiIon usually takes 5 to 6 complete discharge/charge cycles before it really breaks in ? I replaced 2 years ago the battery on my T3 with a higher-capacity version, the exact same thing happened there. It took 5 cycles before I really saw the full power

    I have disconnected my device from the charger yesterday morning at 08h00. Now (over 40 hours later) I still see 22%. GSM was on all the time, I did some pretty heavy BT operations (syncing some +10 Mb podcasts with QuickNews) and did some calls. I am not complaining, certainly not after coming over from a T3
    FWIW I did NOT use the camera. That is a test I will do over the weekend to confirm (from my side) if this is really a problem or not.
    Digital since 1980, Handheld since 2001
    M105, TE, T3.5, Treo 680, Treo 500v
    Happy Datebk6.1 user
  13. #13  
    Quote Originally Posted by mslade View Post
    No one has provided any data that the 12 day standby spec can be confirmed under any conditions.

    The talktime data that I have seen using Batterygraph approximately confirms the specification. Therefore I conclude that it is not a battery capacity problem.

    Since turning the Treo 680 phone off and no use at at should give much longer than 12 days (Bluetooth off, etc.) but apparently does not, it is unlikely that the problem is understood to be an undercapacity battery. It might be a battery that just doesn't hold charge. (I've had mine replace once. Same problem.)
    The Palm OS is still running even with the phone off, BT off, the display off, etc. The only way to totally shut it down is to remove the battery. ALL PALMS ACT THIS WAY. Therefore, to get better than the 'standby time', you would simply remove the battery and measure the free space discharge rate of the battery. I'm not sure that's a useful experiment though.

    We just need to find someone willing to not use their treo for a few days to a week (after doing a total backup) so we can do a true measurement of 'standby time'. Somehow, I don't think we'll find a volunteer there either.

    P.S. Physics is Phun!
  14. ancker's Avatar
    Posts
    52 Posts
    Global Posts
    53 Global Posts
    #14  
    What I wonder is what kind of results people are seeing with the 2400mAh battery Seidio has put out. I haven't seen any posts about it.

    Some conclusions could be made from the data:
    2400mAh lasting roughly 2 times longer than 1200mAh: Treo680 Firmware/software issue
    2400mAh lasting 4hours/300hours: maybe shoddy 1200mAh batteries and possibly Treo680 firmware/software issue
    2400maH lasting 8hours/600hours: shoddy 1200mAh batteries

    /ancker
    Last edited by ancker; 01/05/2007 at 04:12 PM. Reason: fixed typo
  15. mslade's Avatar
    Posts
    16 Posts
    Global Posts
    18 Global Posts
       #15  
    Further research indicates that at least part of the problem is in the battery calibration hardware/firmware - http://forum.brighthand.com/showthread.php?t=235148 .

    I have tried this process and found using BatteryGraph that my initial battery voltage has gone from 3960 MV to 4200MV.

    Recharging and leaving the 680 in standby mode overnight (14 hours) shows voltage down to 4080mv and 82%.

    BatteryGraph also shows that the unit was on for 1 hour 22 minutes during this period during the time it sat on the desk.

    Clearly this being on is causing a drain. BT is off, etc. in my Cingular unit. This version does not seem to have a method of manual networking.
  16. #16  
    Quote Originally Posted by mslade View Post
    Further research indicates that at least part of the problem is in the battery calibration hardware/firmware - http://forum.brighthand.com/showthread.php?t=235148 .
    So is it firmware....
    Quote Originally Posted by mslade View Post
    BatteryGraph also shows that the unit was on for 1 hour 22 minutes during this period during the time it sat on the desk.

    Clearly this being on is causing a drain. BT is off, etc. in my Cingular unit. This version does not seem to have a method of manual networking.
    ...or software??
  17. #17  
    I'm not disputing or challenging anything presented here:
    but all i know is I had 2 Treo 650s, the first was a pre-release version that I wore out and replaced.
    Battery life was never an issue - never went thru a charge-cycle process - used it, ran down over a few days with btooth always on, email always on, screen at full brightness, charged again. Never seemed to need 'training'.
    Granted this battery may be smaller, and I am on my second battery, the usage is not even close. Cannot even make it thru a half day.
    We should not have to go thru all this, it should just work.
    The 680 goes back today. Should not have to be this hard.
    This is not a hobby, it is an important business tool to me.
    Will the new 750 be any better? Same battery? One can only hope.
  18. mslade's Avatar
    Posts
    16 Posts
    Global Posts
    18 Global Posts
       #18  
    I have done an additional test overnight.

    This time the phone was off - not standby, off.

    Using Batterygraph, the battery discharged barely 2-3% over a 17 hour period, suggesting several weeks before a charge was needed.

    This is consistent with what I would have gotten with my Visor PDA.

    My conclusion is that the major problem is with either the phone electronics or firmware or some software ( networking, email) that uses the phone, even in standby.

    This is supported by the previous night's test where BatteryGraph showed that the Treo 680 was on for a total of about two hours overnight when in standby mode even though it was never touched by me.

    My sense from reading some of the other forum comments is that this is not a Cingular Treo 680 problem but true of all 680s. Can anyone confirm this?

    Are the newer Treo 680s any different in their operation?

    Does Palm have a fix yet?

    Anyone else have any data that would shed more light on the cause?

    thanks
  19. #19  
    As far as I understood standby is exactly what you have tested: PDA functionality only with phone off.
    When phone is on the device will connect to the network once in a while to get a fix on the radio antennas, to see if it does not need to switch to another antenna, etc. etc.

    I did some tests as well and can get almost 50 hours with light use if I do NOT use the camera.
    Digital since 1980, Handheld since 2001
    M105, TE, T3.5, Treo 680, Treo 500v
    Happy Datebk6.1 user
  20. mrjoec's Avatar
    Posts
    369 Posts
    Global Posts
    384 Global Posts
    #20  
    I agree that Palm definitely considers 'standby' to be 'phone turned off'. That's the only way that the 300 hour claim can be even close to reality.

    However, calling that state 'standby' is dubious, to say the least. Standby means on, but sitting idle, ready to be used. When my guitar amp is in standby mode, power is coursing through it, the tubes are warmed up, and a quick flick of a switch causes instant hard rock euphoria.

    Any other phone in the industry wouldn't use turning the phone off as standby mode. If that were the case, Nokia could claim to have a standby battery life of infinity.
    mrjoec
    www.joecieplinski.com
Page 1 of 4 1234 LastLast

Posting Permissions